Exploring the Effects of Simulation-Based Instruction for Neural Networks

This study investigates the effectiveness of simulation-based instruction in helping students understand concepts of Neural Networks, explicitly focusing on topics related to Deep Learning. The proposed simulation-based instruction aims to help students learn abstract concepts and foster their ability to apply deep learning knowledge to solve problems flexibly. A quasi-experimental study is conducted to compare simulation-based instruction with traditional instruction. The research findings reveal that the proposed simulation-based instruction can help students with moderate to low prior knowledge comprehend deep learning concepts. Additionally, students exhibited higher creativity in terms of “sensitivity” and “originality” in the experimental group than in the control group. Students in the experimental group also had more positive perceptions of creativity in information technology. Through our instructional guidance and simulation tools, students can clarify their concepts and understand how to apply them in different problem scenarios.